Policy Exchange has just published a “research note” purporting to show that the tax on cigarettes in the UK should be increased, and that “that every single cigarette smoked costs the country money – 6.5 pence each time someone lights up.”
If you read the paper [pdf], you will find it is an astonishingly dodgy dossier. Here is how the figure is made up:
Taxation of tobacco contributes £10 billion to HM Treasury annually; however, we calculate that the costs to society from smoking are much greater at £13.74 billion. Every cigarette smoked is costing us money. These societal costs comprise not only the cost of treating smokers on the NHS (£2.7 billion) but also the loss in productivity from smoking breaks (£2.9 billion) and increased absenteeism (£2.5 billion); the cost of cleaning up cigarette butts (£342 million); the cost of smoking related house fires (£507 million), and also the loss in economic output from the deaths of smokers (£4.1 billion) and passive smokers (£713 million).
The notion of “cost to society” is a pretty weird one.
Leave that aside for a moment. Add up costs and revenues to the state, which might be one semi-logical way of determining whether the smoking in some sense “runs a deficit”, and using Policy Exchange’s own figures you get a big surplus for the Treasury. Even if you assume all house fire costs are borne by the state and not partially by insurers and householders, and there are no errors in the headline figures, then you can only get to £3,549 million. (Have you noticed how public policy research generally involves implausible numbers of significant digits, and at the same time utter absence of error estimates?) On that basis smokers are contributing roughly £6Bn annually towards public spending.
But what are we to make of the suggestion that counting “lost output” is meaningful? To my mind the idea that an economic aggregate represents a collective wealth that may be politically attributed and redistributed is repulsive even if it is coherent (which I doubt). The state’s royal We, which Policy Echange is channelling here, may in turn choose to impersonate you and me and everyone else, but it only controls the taxed margin of other’s outputs. Output and taxation are apples and oranges. It is meaningless to add them together. Unless you want (or deserve) a punch.
And even were it not meaningless, there’s an accounting fraud here. If you count output putatively lost to smoking, then you must also count the gains. There is the output of the tobacco industry, distribution and retailing in the UK to consider. Imperial Tobacco alone had a gross profit for the year ending September 2009 of approximately £5.3 billion. The CTC industry consists of tens of thousands of small shops. Honest research, however dubious its theoretical basis, would attempt to estimate the value-added, too. It would also be clear – without referring to a paper cited in the footnotes we cannot tell whether the cost-of-illness measure used in determining those “lost outputs” also includes the gains to third parties in pensions unpaid and public services unused by people dying early. If you are going to add apples and oranges, you should also tell us explicitly whether you have subtracted pears.
But what set me off on this chase was actually just one of those headline figures. Most of the margin of costs over gains in this strange sum is covered by the £2.9 billion allocated to the “output lost to cigarette breaks”. How do they know? “[A] number of studies have investigated workers taking breaks in order to smoke, and have tried to quantify this time at between £915 million and £3.2 billion per annum.” Hm.
Read through to p13, and you discover that the number of studies was… two. Er, no. It was one… Or some sort of strange interpolative hybrid… I cannot decide. Make your own mind up:
McGuire et al. estimated that £915 million annually is lost on the basis that average smokers spend tenminutes a day smoking, while light smokers and part-time workers would use approximately half of this
time. The Royal College of Physicians (RCP) used similar initial assumptions on average smoking time to
calculate that some £2.6 billion would be saved through the introduction of smoke-free legislation. Using
McGuire’s estimates of 5.2 million working smokers, with the RCP’s estimates of ten minutes a day smoking
reveals an intermediary figure of £2.9 billion.
I think that is ‘intermediary’ in the sense that a magician is an intermediary between a rabbit and a hat.
However they get there, if someone thinks that cigarette breaks ought to be a determining factor in public policy, rather than a matter for negotiation between employer and employee, then I suggest that it would be a good idea if they are kept as far as possible from the levers of power. This lot are said to be influential on the presumptively incoming Cameron team. Oh dear.